Method, device, and computer program product for rendering image

Abstract

Embodiments of the present disclosure relate to a method, a device, and a computer program product for rendering an image. The method includes determining, based on a coordinate value of a target point in a target scenario and a viewing direction for the target point, a rendering parameter for the target point. The method further includes adjusting the rendering parameter based on the viewing direction and by performing an upsampling operation on the rendering parameter. The method further includes rendering an image for the target scenario based on the adjusted rendering parameter. By means of the method, resources required by processing of image data are reduced; the image data processing speed is increased; and fast rendering of a high-resolution image can be achieved.

What is claimed is: 
     
         1 . A method, comprising:
 determining, based on a coordinate value of a target point in a target scenario and a viewing direction for the target point, a rendering parameter for the target point;   adjusting the rendering parameter based on the viewing direction and by performing an upsampling operation on the rendering parameter; and   rendering an image for the target scenario based on the adjusted rendering parameter.   
     
     
         2 . The method according to  claim 1 , wherein the rendering parameter comprises a color parameter, and the determining a rendering parameter for the target point comprises:
 determining a multidimensional data representation of a color parameter for the target point based on the coordinate value of the target point;   determining, based on the viewing direction, a weight set related to the multidimensional data representation; and   determining the color parameter for the target point based on the multidimensional data representation and the weight set.   
     
     
         3 . The method according to  claim 2 , wherein the rendering parameter further comprises a volume density; the volume density is used for indicating a transparency of the target point; and the determining a rendering parameter for the target point further comprises:
 determining a volume density for the target point based on the coordinate value of the target point; and   determining the rendering parameter based on the volume density and the color parameter.   
     
     
         4 . The method according to  claim 1 , wherein the determining a rendering parameter for the target point comprises:
 determining the rendering parameter using a first fully connected neural network model.   
     
     
         5 . The method according to  claim 4 , further comprising:
 determining a volume corresponding to the target scenario; and   dividing the volume into a plurality of sub-volumes, wherein the sub-volumes in the plurality of sub-volumes comprise the target point.   
     
     
         6 . The method according to  claim 5 , further comprising:
 performing parallel processing on points in the plurality of sub-volumes using a plurality of fully connected neural network models, the plurality of fully connected neural network models comprising the first fully connected neural network model.   
     
     
         7 . The method according to  claim 1 , wherein the adjusting the rendering parameter comprises:
 generating a two-dimensional data representation for the rendering parameter and the viewing direction based on the rendering parameter and the viewing direction; and   adjusting the rendering parameter based on the two-dimensional data representation and by upsampling the two-dimensional data representation.   
     
     
         8 . The method according to  claim 7 , wherein the generating a two-dimensional data representation for the rendering parameter and the viewing direction comprises:
 rearranging the value of each parameter in the rendering parameter to generate the corresponding two-dimensional data representation; and   determining the two-dimensional data representation for the viewing direction based on the viewing direction and a predetermined matrix.   
     
     
         9 . The method according to  claim 7 , wherein the adjusting the rendering parameter based on the two-dimensional data representation and by upsampling the two-dimensional data representation comprises:
 performing feature extraction on the two-dimensional data representation to obtain a first feature representation;   upsampling the first feature representation to obtain an extended feature representation;   performing feature extraction on the extended feature representation to obtain a second feature representation;   upsampling the two-dimensional data representation to obtain an extended two-dimensional data representation; and   adjusting the rendering parameter based on the second feature representation and the extended two-dimensional data representation.   
     
     
         10 . The method according to  claim 9 , wherein the upsampling the first feature representation to obtain an extended feature representation comprises:
 interpolating the first feature representation to obtain the extended feature representation.   
     
     
         11 . The method according to  claim 1 , wherein the rendering an image for the target scenario comprises:
 rendering the image for the target scenario based on the viewing direction and the adjusted rendering parameter.   
     
     
         12 . An electronic device, comprising:
 at least one processor; and   a memory coupled to the at least one processor and having instructions stored thereon, wherein the instructions, when executed by the at least one processor, cause the electronic device to perform actions comprising:   determining, based on a coordinate value of a target point in a target scenario and a viewing direction for the target point, a rendering parameter for the target point;   adjusting the rendering parameter based on the viewing direction and by performing an upsampling operation on the rendering parameter; and   rendering an image for the target scenario based on the adjusted rendering parameter.   
     
     
         13 . The electronic device according to  claim 12 , wherein the rendering parameter comprises a color parameter, and the determining a rendering parameter for the target point comprises:
 determining a multidimensional data representation of a color parameter for the target point based on the coordinate value of the target point;   determining, based on the viewing direction, a weight set related to the multidimensional data representation; and   determining the color parameter for the target point based on the multidimensional data representation and the weight set.   
     
     
         14 . The electronic device according to  claim 13 , wherein the rendering parameter further comprises a volume density; the volume density is used for indicating a transparency of the target point; and the determining a rendering parameter for the target point further comprises:
 determining a volume density for the target point based on the coordinate value of the target point; and   determining the rendering parameter based on the volume density and the color parameter.   
     
     
         15 . The electronic device according to  claim 12 , wherein the determining a rendering parameter for the target point comprises:
 determining the rendering parameter using a first fully connected neural network model.   
     
     
         16 . The electronic device according to  claim 15 , further comprising:
 determining a volume corresponding to the target scenario; and   dividing the volume into a plurality of sub-volumes, wherein the sub-volumes in the plurality of sub-volumes comprise the target point.   
     
     
         17 . The electronic device according to  claim 16 , wherein the actions further comprise:
 performing parallel processing on points in the plurality of sub-volumes using a plurality of fully connected neural network models, the plurality of fully connected neural network models comprising the first fully connected neural network model.   
     
     
         18 . The electronic device according to  claim 12 , wherein the adjusting the rendering parameter comprises:
 generating a two-dimensional data representation for the rendering parameter and the viewing direction based on the rendering parameter and the viewing direction; and   adjusting the rendering parameter based on the two-dimensional data representation and by upsampling the two-dimensional data representation.   
     
     
         19 . The electronic device according to  claim 18 , wherein the generating a two-dimensional data representation for the rendering parameter and the viewing direction comprises:
 rearranging the value of each parameter in the rendering parameter to generate the corresponding two-dimensional data representation; and   determining the two-dimensional data representation for the viewing direction based on the viewing direction and a predetermined matrix.   
     
     
         20 . A computer program product that is tangibly stored on a non-transitory computer-readable medium and comprises machine-executable instructions, wherein the machine-executable instructions, when executed by a machine, cause the machine to perform a method, the method comprising:
 determining, based on a coordinate value of a target point in a target scenario and a viewing direction for the target point, a rendering parameter for the target point;   adjusting the rendering parameter based on the viewing direction and by performing an upsampling operation on the rendering parameter; and   rendering an image for the target scenario based on the adjusted rendering parameter.